Kinetic study of the condensation of salicylaldehyde with diethyl malonate in a catalyzed nonpolar solvent by secondary amines

Article abstract of DOI:10.1002/kin.20429

The kinetics of condensation between salicylaldehyde and diethylmalonate in a toluene solution in the presence of secondary amines (i.e., piperidine and 4-piperidinopiperidine) as catalysts was investigated. It was found that the reaction proceeds via the Knoevenagel mechanism, and the kinetic model was numerically verified. copy;2009 Wiley Periodicals.

Full text of DOI:10.1002/kin.20429

Kinetic Study of the
Condensation of
Salicylaldehyde with
Diethyl Malonate in a
Nonpolar Solvent Catalyzed
by Secondary Amines
SZCZEPAN BEDNARZ,1 DARIUSZ BOGDAL2
1Department of Engineering and Machinery for Food Industry, University of Agricultural in Krakow,
ul. Balicka 122, 30-149 Krakow, Poland
2Faculty of Chemical Engineering and Technology, Krakow University of Technology,
ul. Warszawska 24, 31-155 Krakow, Poland
Received 24 June 2008; revised 24 February 2009; accepted 3 March 2009
DOI 10.1002/kin.20429
Published online in Wiley InterScience (www.interscience.wiley.com).
ABSTRACT: The kinetics of condensation between salicylaldehyde and diethylmalonate
in
a toluene solution in the presence of secondary amines (i.e., piperidine and
4-piperidinopiperidine) as catalysts was investigated. It was found that the reaction proceeds
via the Knoevenagel mechanism, and the kinetic model was numerically verified. ꢀ 2009 Wiley
C
Periodicals, Inc. Int J Chem Kinet 41: 589–598, 2009
ing on the kind of a catalyst were proposed [1]. The
Hann and Lapworth mechanism assumes the forma-
tion of the β-hydroxy intermediate as a product of the
addition of a carbanion, which is produced in the reac-
tion of an “active methylene compound” and base, to
an aldehyde [2]. In turn, the Knoevenagel mechanism
is proposed to proceeds via an active amino compound
(e.g., iminium ion, aminal) that is formed from an alde-
hyde and amine [3].
INTRODUCTION
The Knoevenagel condensation is generally defined as
the reaction between an aldehyde and compounds with
“active methylene groups” in the presence of organic/
inorganic bases, ammonia, or their salts. This wide
definition includes a number of reactions with differ-
ent reactivities of substrates, kind of catalyst, as well as
polarity of solvents. Finally, two mechanisms depend-
The Knoevenagel condensation is one of the im-
portant methods for the coumarins synthesis [4]. Re-
cently, the influence of microwave irradiation on the
kinetics of such reactions was investigated and an in-
crease of the reaction rate under microwave conditions
Correspondence to: Szczepan Bednarz; e-mail: sbednarz@
ar.krakow.pl.
Figures 1–4 are available as supporting information in the online
issue at www.interscience.wiley.com.
c
ꢀ
2009 Wiley Periodicals, Inc.

590
BEDNARZ AND BOGDAL
was observed [4,5]. To explain the origin of such an
effect, the mechanism of these kinds of reactions was
investigated.
Aminals Preparation
2.55 g pipridine (30 mmol) was dropped to 1.22 g sal-
icylaldehyde (10 mmol) and vigorously mixed. Next,
the mixture was kept in a refrigerator for 5 days. The
crude solid product was purified by twice recrystaliza-
tion from hexane, yielding a white solid, m.p. 87◦C;
H NMR 300 MHz, CDCl3): 12,0 (s, 1H), 7.3–7.1 (m,
1H), 6.9–6.7 (m, 3H), 3.75 (s, 1H), 2.5 (s, 8H), 1.5 (d,
12H).
5.04 g 4-piperidinopiperidine (30 mmol) was dis-
solved in minimal amount of dichloromethane, and
the solution was dropped to 1.22 g salicylaldehyde
(10 mmol) and vigorously mixed. Next, the mixture
was put in the refrigerator for 5 days. The crude solid
product was purified by twice recrystallization from
hexane, yielding a white solid, m.p. 140◦C; H NMR
(300 MHz, CDCl3): 11.63 (s, 1H), 7.26–7.14 (m, 1H),
6.79–6.75 (m, 3H), 3.76 (s, 1H), 3,00 (d, J = 12 Hz,
2H), 2.56–2.18 (br, 14H), 1.74–1.57 (br, 22H).
EXPERIMENTAL
Materials
Piperidine was heated with solid potassium hy-
droxide for 5 h under reflux condenser. Then it
was distilled and stored in a closed vessel. 4-
Piperidinopiperidine (98%; Sigma-Aldrich, St. Louis,
MO, USA) was recrystallized from hexane. Salicy-
laldehyde (>99%, Fluka, Steinheim, Germany), di-
ethylmalonate (>99%; Fluka), morpholine (>98%;
Fluka), diazabicyclo[2.2.2]octane—DABCO (98%;
Sigma-Aldrich), Et3N (>98%; POCH) and toluene
(POCH) were used as received.
General Methods
Numerical Calculations
Melting points, measured on an Electrothermal IA9200
microscope plate, are uncorrected. The progress of re-
actions was monitored by both a gas chromatograph
(GC) HewlettPackard 5890 coupled with a mass de-
tection (MS) HewlettPackard 5971 and a gas chro-
matograph Agilent 6850 with a flame ionization detec-
tor (FID). Both chromatographs were equipped with
HP-1 columns. 1H NMR spectra were recorded with a
Bruker AVANCE-300 spectrometer. FT-Raman spec-
tra were obtained using an EZRaman-M spectrometer,
using 670-nm excitation and 200-mW power laser.
On the basis of the postulated reaction mechanism,
a system of differential equations describing changes
of concentration each chemical compounds was made.
Rate constants were estimated by the least-square fit-
ting method (based on the Levenberg–Marquardt algo-
rithm) using computational program Dynafit [6]. Min-
imized function was as follows:
Nj
ꢀ ꢀ
M
2
ˆ
([C]j,i − [C]j,i
)
j=1 i=1
Kinetic Investigations
where M is the number of runs with different ini-
tial substrate concentrations [A]j,1,[M]j,1; Nj is the
number of measurements in j run (thus j=1 Nj =
N express the total number of data points); [C]j,i
is the product concentration, expressed as a mean
An appropriate amount of substrates, catalyst (piperi-
dine or 4-piperidinopiperidine), and naphthalene (in-
ternal standard) were dissolved in a toluene and placed
in a flask, closed and kept at room temperature until
more than 90% substrate conversion was obtained (it
took about 10 days). At appropriate time intervals, a
sample of the mixture was withdrawn and diluted in
acetone (it was found that aminals easily decompose
to starting materials in the acetone solution). Reaction
progress was monitored by means of GC-FID, using
naphthalene as an internal standard.
ꢁ
R
value based on the substrate consumption: [C]j,i
=
ˆ
([A]j,1 − [A]j,i + [M]j,1 − [M]j,i)/2; [C]j,i is the es-
timated product concentration; [A]j,i,[M]j,i are mea-
sured substrate concentrations.
The quality of fitting was expressed as sum of
squares and a relative error of rate constants estima-
tions. Both parameters were calculated by means of
Dynafit.
Catalytic Tests
RESULTS AND DISCUSSION
Nonkinetic Consideration
The mixture of 10 mL of toluene, 5 mmol of catalyst
(DABCO, Et3N, or morpholine), 0.2 mL (20 mmol)
salicylaldehyde, 0.4 mL (25 mmol) diethylmalonate
was kept at 70◦C for a 4 h. The presence of reaction
product by GC-MS was found.
The aim of the presented work was to determine the
condensation mechanism of salicylaldehyde (A) and
International Journal of Chemical Kinetics DOI 10.1002/kin

KINETIC STUDY OF THE CONDENSATION OF SALICYLALDEHYDE WITH DIETHYL MALONATE
591
Scheme 1 The reaction of salicylaldehyde (A) and diethylmalonate (M) in the presence of a secondary amine (P).
Scheme 2 Possible paths of 2,2-diethoxycarbonyl-1-phenyl-ethenol-1 decomposition.
diethylmalonate (M) catalyzed by a secondary amine
(P) in a nonpolar solvent (Scheme 1).
In turn, the condensation of aromatic aldehydes (e.g.,
benzaldehyde) with various compounds with active
methylene groups of different acidities in the absence
of any catalyst followed the Hann–Lapworth mecha-
nism, and the ionization of a compound with an active
methylene group was the rate-determining step [10a–
10d].
The investigated reaction occurs via two main
steps, i.e., the Knoevenagel condensation that af-
fords a benzylidene derivative (B) and then lactoniza-
tion of the derivative B that yields the final product
3-ethoxycarbonylcoumarin (C). The second step seems
to be irreversible since the analysis of a solution of the
coumarin C in EtOH even after heating was not in-
dicating a retroreaction toward B and then A and M.
Moreover, the last step (lactonization) is a fast one and
the traces of unsaturated derivative B was detected by
means of chromatographic methods (GC/MS, HPLC).
Therefore, it was possible to simplify the investigated
reaction and consider only the first step, i.e., the Kno-
evenagel condensation.
The kinetics of the Knoevenagel condensation of
benzaldehyde with malonic ester in the presence of
secondary amines as a catalyst has already been inves-
tigated [7,8]. It was found that the β-hydroxy product
was very resistant to dehydratation and easily decom-
posed to the starting materials (Scheme 2). Thus, the fi-
nal product of the reaction is formed from the β-amino
compounds, which supported the Knoevenagel mech-
anism. A similar mechanism for the condensation of
methyl arylsulfinylacetate with various aldehydes was
proposed by Tanikaga et al. [9]. A further study showed
that the deamination of β-amino compound is the rate-
limiting step (Scheme 3), which can be accelerated in
the presence of H+ (e.g., as the protonated amine) [7].
According to the Hann and Lapworth mechanism,
the catalytic activity of tertiary amines increases when
the basicity of an amine rises and does not depend on its
structure. On the other hand, the Knoevenagel mecha-
nism requires only secondary or primary amines as a
catalyst, and tertiary amines do not catalyze the con-
densation because they cannot form amino intermedi-
ates in the reaction with an aldehyde. To decide which
of two presented mechanisms is more probable in the
reaction of salicylaldehyde (A) and diethyl malonate
(M), catalytic tests were performed in the presence of
secondary as well as tertiary amines (Table I).
Based on the results shown in Table I, the condensa-
tion of salicylaldehyde (A) and diethyl malonate (M)
in a toluene solution proceeds via the Knoevenagel
mechanism rather than the Hann–Lapworth mecha-
nism. Tertiary amines (i.e., diazabicyclo[2.2.2]octane
(DABCO) and Et3N) did not catalyze the reaction, but
secondary amines (i.e., morpholine and piperidine) that
posses similar basicity to the tertiary amines catalyzed
the reaction.
Furthermore, in the case of the Knoevenagel con-
densation catalyzed by secondary amines, the reaction
International Journal of Chemical Kinetics DOI 10.1002/kin

592
BEDNARZ AND BOGDAL
Scheme 3 Acid-catalyzed deamination of β-amino compound.
Table I Correlation between Catalytic Activity of the
Selected Secondary and Tertiary Amines and Their
Basicity
the other hand, it is well known that iminium cations
like AP+ are intermediates in the Mannich condensa-
tion, in which the cations are generated from formalde-
hyde and an amine [13].
Catalytic Activity
1
In our study, H NMR analysis of the equimolar
Base
pKb
Yield of Coumarin (%)
mixture of salicylaldehyde and piperidine showed that
the aminal AP2 was readily formed (Scheme 5) and
the reaction reached the equilibrium stage quickly (see
Supporting Information Fig. 1). Aminals from sali-
cylaldehyde and other secondary amines can also be
easily prepared (see the Experimental section).
DABCO
Et3N
Morpholine
Piperidine
5.8/9.8a
3.2
5.3
0
0
15
53
2.9
apK1 and pK2 values are taken from [11].
1
In addition, H NMR analysis does not give in-
formation about formation of other compounds such
as the hemaminal AP or iminium ion AP+. Recently,
the kinetics of the Knoevenagel condensation of sali-
cylaldehyde with benzyl acetoacetate in the presence
of piperidine by means of Raman spectroscopy was
investigated [14], and the formation of benzylidene-
like product was observed. However, in the case of
condensation of salicylaldehyde and diethylmalonate,
the aminal AP2 was the only intermediate detected by
means of the method.
Eventually, only the aminal AP2 as intermediate
was detected and prepared in our investigation. The
remaining compounds (intermediates) were not de-
tected by means of chromatographic methods (HPLC,
GC/MS) and Raman spectroscopy. It indicates that
most of preliminary reactions, i.e., (1), (2), (3), (6), and
(7) (Scheme 4) are equilibrium reactions, fast in both
directions, and, for this reason, the concentration of
intermediates is under the detection limit of analytical
methods used.
proceeds via intermediates like an aminal (AP2) or
an imminium salt (AP+), which is considerably more
electrophilic than the aldehyde [1]. In accordance with
that a mechanism is outlined in Scheme 4.
In addition to the main reactions, side reactions can
occur, i.e., carbanion generation via deprotonation of
the active methylene groups by a base (amine) and
addition of the carbanion to the aldehyde to give the
β-hydroxy compound as well as the acid–base reaction
between salicylaldehyde and the amine catalyst.
Reaction Intermediates
On the basis of detailed kinetic investigations on
the aldolic stage of the Knoevenagel condensation of
benzaldehyde and diethyl malonate, Kinastowski and
Mroczyk [7,8] suggested that the condensation via an
intermediate β-amino compound is most probable. The
β-amino compound was formed as the product of reac-
tions of diethylmalonate with two key intermediates,
i.e. aminal or/and iminium cation. The synthesis of
aminals with similar structure to the aminal AP2 from
benzaldehyde and piperidine was already described by
Patai et al. [10e]. Later, Tanaka et al. proved the struc-
ture of these aminals using NMR spectroscopy [12]. On
Kinetics Models
As can be seen in Scheme 4, the investigated reaction
is complex. Moreover, there is a lack of strict infor-
mation on both formation and concentration of each
International Journal of Chemical Kinetics DOI 10.1002/kin

KINETIC STUDY OF THE CONDENSATION OF SALICYLALDEHYDE WITH DIETHYL MALONATE
593
Scheme 4 Proposed course of the Knoevenagel condensation of salicylaldehyde with diethylmalonate, catalyzed by secondary
amine in a nonpolar solvent.
intermediate (i.e., AP, AP+, B). Nevertheless, kinetic
alyst has also another interesting feature—it possesses
an additional tertiary amine group.
modeling was chosen to prove the reaction mechanism.
It was found that the contact of piperidine with H2O
and CO2 from the air causes generation of piperidine
cation, which influences the kinetics of the Knoeve-
nagel condensation [15]. However, in the investigated
system there is a phenolic derivative, salicylaldehyde,
which can act as a proton donor and protonates the
amine catalyst, thus the influence of air may be ne-
glected. Moreover, we also chose to use as a catalyst 4-
piperidinopiperidine, which is a secondary amine that
is more stable, less hygroscopic, easier to purify, and
simultaneously soluble in nonpolar solvents. This cat-
According to the above considerations, it is possible
to define three reaction paths (Scheme 4), i.e. 1-2-5-
7-8,∗ 1-2-4-6-7-8, and 1-3-6-7-8, and for each path a
numeric model (a system of differential equations) was
created: K-1, K-2, K-3, respectively (Tables II and III).
Then, the rate constants were numerically estimated
for each kinetic model for the reactions catalyzed by
both 4-piperidinopiperidine and piperidine. Next, the
complex equation system was simplified by accepting
reducing assumptions and model parameters for each
approximation (A, B, C, D, and E) were determined
(Tables III–VI).
Approximation A (K1A, K2A, K3A)
1. Hydronium ion concentration is constant and
thus k4ꢁ = [H+]k4, k3ꢁ = [H+]k3; however, it is
difficult to control the pH in such nonaqueous
systems.
∗For reaction numbers see Scheme 4. Rate constants are num-
bered in the same manner.
Scheme 5 Aminal formation from salicylaldehyde and
piperidine.
International Journal of Chemical Kinetics DOI 10.1002/kin

594
BEDNARZ AND BOGDAL
Table II Kinetic Models. Approximations A and B
Model Name
Scheme
System of Differential Equations
= −k1[A][P] + k−1[AP]
d[A]
dt
d[P]
= −k1[A][P] + k−1[AP] − k2[P][AP] + k−2[AP2][W]
+ 2k5[AP2][M]
dt
k1
d[AP ]
→
K-1A
K-1B
A + P
AP
= k1[A][P] − k−1[AP] − k2[P][AP] + k−2[AP2][W]
= k2[P][AP] − k−2[AP2][W] − k5[AP2][M]
k
dt
d[AP2]
−1
k2
→
AP + P
AP2 + W
k
dt
−2
k5
d[W]
→
AP2 + M 2P + C + E
= k2[P][AP] − k−2[AP2][W]
dt
d[M]
= −k5[AP2][M]
dt
d[C]
d[E]
=
= k5[AP2][M]
dt
dt
d[A]
= −k1[A][P] + k−1[AP]
dt
d[P]
= −k1[A][P] + k−1[AP] − k2[P][AP] + k−2[AP2][W]
+ k4ꢁ [AP2] − k−ꢁ 4[P][AP+] + k5[AP+][M]
dt
k1
d[AP]
↔
A + P
AP
= k1[A][P] − k−1[AP] − k2[AP][P] + k−2[AP2][W]
= k2[P][AP] − k−2[AP2][W] − k4ꢁ [AP2] + k−ꢁ 4[P][AP+]
k
dt
d[AP2]
−1
k2
↔
K-2A
K-2B
AP + P
AP2 + W
k
k4ꢁ
dt
d[W]
−2
↔
AP2
AP+ + P
= k2[P][AP] − k−2[AP2][W]
k−ꢁ 4
dt
d[AP+]
k6
AP+ + M 2P + C + E
= k4ꢁ [AP2] − k−ꢁ 4[P][AP+] − k6[AP+][M]
= −k6[AP+][M]
→
dt
d[M]
dt
d[C]
d[E]
=
= k6[AP+][M]
dt
d[A]
dt
= −k1[A][P] + k−1[AP]
= −k1[A][P] + k−1[AP] + k6[AP+][M]
dt
k1
d[P]
↔
K-3A
K-3A
A + P
AP
k
dt
−1
k3ꢁ
d[AP]
AP
AP + W
= k1[A][P] − k−1[AP] − k3ꢁ [AP] + k5ꢁ [AP+][W]
+
↔
k−ꢁ 3
dt
d[AP+]
= k3ꢁ [AP] − k−ꢁ 3[AP+][W]
= −k6[AP+][M]
dt
d[M]
k6
AP+ + M ↔ P + C + E
dt
d[C]
d[E]
=
= k6[AP+][M]
dt
dt
2. β-Amino intermediate (APM) formation (5) and
(6) is an irreversible step and thus k5ꢂk
Approximation B (K1B, K2B, K3B).
,
−5
k6ꢂk
.
−6
1. On the basis of additional numerical simulations,
3. Amine elimination (7) and lactonization (8) are
fast steps.
we found that adjusting values of k1/k ≈ 1 and
−1
k1 ≈ k ≈ 1000 model fitting is the best. How-
−1
As can be seen (Table IV), this model does not fit
to experimental data and next reductions should be
applied.
ever, both the ratio and the absolute rate constant
values can vary and they do not influence both
model fitting and values of the other estimated
International Journal of Chemical Kinetics DOI 10.1002/kin

KINETIC STUDY OF THE CONDENSATION OF SALICYLALDEHYDE WITH DIETHYL MALONATE
595
Table III Kinetic Models: Approximations C, D, and E
Model Name
K-1C
Scheme
System of Differential Equations
kr
d[A]
A + 2P
AP2 + W
= −kr [A][P]2 + k−r [AP2][W]
= −2kr [A][P]2 + 2k−r [AP2][W] + 2k5[AP2][M]
= kr [A][P]2 − k−r [AP2][W]
↔
k
dt
d[P]
−r
k5
→
AP2 + M
C + 2P + E
dt
d[W]
dt
d[M]
= −k5[AP2][M]
dt
d[C]
d[E]
=
= k5[AP2][M]
= −kr [A][P]2 + k−r [AP2][W]
= −2kr [A][P]2 + 2k−r [AP2][W] + k5[AP2][M] + k7[APM]
=kr [A][P]2 − k−r [AP2][W] − k5[AP2][M]
dt
d[A]
dt
kr
↔
K-1D
A + 2P
AP2 + W
k
dt
d[P]
−r
k5
↔
AP2 + M
APM + P
dt
d[AP2]
k7
→
APM
C + P + E
dt
d[W]
= kr [A][P]2 − k−r [AP2][W]
dt
d[M]
= −k5[AP2][M]
dt
d[APM]
= k5[AP2][M] − k7[APM]
dt
d[C]
d[E]
=
= k7[APM]
= −kr [A][P]2 + k−r [AP2][W]
= −2kr [A][P]2 + 2k−r [AP2][W] + k5[AP2][M] + k7[APM]
= kr [A][P]2 − k−r [AP2][W] − k5[AP2][M]
dt
d[A]
dt
kr
↔
A + 2P AP2 + W
−r
K-1E
k
dt
d[P]
k5
→
AP2 + M APM + P
dt
d[AP2]
k7
→
APM B + P
dt
d[W]
k8
→
B
C + E
= kr [A][P]2 − k−r [AP2][W]
dt
d[M]
= −k5[AP2][M]
dt
d[APM]
= k5[AP2][M] − k7[APM]
dt
d[B]
= k7[APM] − k8[B]
dt
d[C]
d[E]
=
= k8[B]
dt
dt
rate constants. The results of this approximation
(Table V) are quite sufficient, especially in the
case of K-1B model. Therefore, the K-1 model
was evaluated.
from numerical calculations and the experimen-
tal data for different concentrations of salicy-
laldehyde (A) with diethylmalonate (M) are pre-
sented (see Supporting Information Figs. 2 and
3). It can be seen that the reaction catalyzed by
piperidine is faster than 4-piperidinopiperidine
because of higher value of the rate constant of
the second step (k2). Moreover, the equilibrium
(Scheme 5) is established more rapidly for the
piperidine-catalyzed reaction than in the case of
4-piperidinopiperidine but the values of Kr are
similar. These effects could be explained as steric
effects because piperidine molecules are smaller
than 4-piperidinopiperidine and can react faster.
Approximation C (K1C).
1. Hemiaminal formation (1) was omitted, and the
reaction of aminal generation as one step was
considered. As can be seen (Table VI), fitting
of model K-1C is the best while considering the
sum of squares values and relative errors of rate
constants. Next, the comparison of the results
International Journal of Chemical Kinetics DOI 10.1002/kin

596
BEDNARZ AND BOGDAL
Table IV Results of Model Fitting: Approximation A
Kinetic Model
4-Piperidinopiperidine
Piperidine
K-1A
SSR
SD
k1
5 × 10−5
2 × 10−5
7.00 × 10−3
5.00 × 10−3
6.11 × 10−4 110%
2.11 × 10−2 1700%
2.06 × 10−1 1700%
9.17 × 10−4 62%
3.42 × 10−4 16%
1.39 × 10−3 280%
9.44·10−3 940%
4.83 × 10−2 670%
7.50 × 10−4 67%
4.72 × 10−4 24%
2.9 × 10−5
k
−1
k2
k
−2
k5
SSR
SD
k1
K-2A
5.00 × 10−3
4.31 × 10−3 430%
3.89 × 10−3 440%
5.31 × 10−3 77%
4.86 × 10−2 950%
1.74 × 10−1 520%
3.94 × 10−2 880%
3.61 × 10−4 42%
5 × 10−5
k
−1
k2
Not estimated
k
−2
k4ꢁ
k−ꢁ 4
k6
K-3A
SSR
SD
k1
7.5 × 10−5
8.00 × 10−3
7.00 × 10−3
3.03 × 10−4 360%
2.97 × 10−3 2500%
2.22 × 10−2 2900%
1.47 × 10−3 2600%
1.36 × 10−4 26%
2.58 × 10−4 1100%
9.06 × 10−4 2500%
4.47 × 10−3 4600%
1.33 × 10−3 5300%
2.11 × 10−4 44%
k
−1
kꢁ3
kꢁ
−3
k6
Table V Results of Model Fitting: Approximation B
Kinetic Model
K-1B
Kinetic Model
4-Piperidinopiperidine
Piperidine
SSR
SD
k2
7 × 10−5
2 × 10−5
8.00 × 10−3
5.00 × 10−3
4.61 × 10−3 17%
1.00 × 10−3 22%
4.72 × 10−4 14%
6 × 10−5
7.36 × 10−3 16%
8.33 × 10−4 18%
5.28 × 10−4 14%
3 × 10−5
k
−2
k5
SSR
SD
k2
K-2B
K-3B
8.00 × 10−3
5.00 × 10−3
4.72 × 10−3 26%
2.94 × 10−2 1100%
2.03 × 10−1 620%
3.78 × 10−2 950%
2.11 × 10−4 30%
8 × 10−5
5.50 × 10−3 21%
1.56 × 10−2 610%
4.39 × 10−2 450%
3.00 × 10−2 390%
3.61 × 10−4 40%
5 × 10−5
k
−2
kꢁ4
kꢁ
−4
k6
SSR
SD
kꢁ3
kꢁ
k6
9.00 × 10−3
7.00 × 10−3
3.17 × 10−4 17%
2.17 × 10−4 23%
1.53 × 10−4 13%
2.33 × 10−4 25%
2.64 × 10−4 27%
2.31 × 10−4 27%
−3
The model K-1C was studied in more detail (models
K-1D and K-1E). But taking into consideration more
elementary steps gave poor results of the models fit-
ting. Since there is no significant change of fitting qual-
ity comparing 4-piperidinopiperidine with piperidine-
catalyzed reaction, it could be postulated that second
N-atom (tertiary) of 4-piperidinopiperidine does not
influence directly the reaction mechanism. This is in
an agreement with the previous results presented in
Table I, in which DABCO and Et3N did not catalyze
the reaction.
It was also interesting to analyze the kinetic results
in the case of a significant excess of one of the sub-
strates. When salicylaldehyde (A) was in excess (i.e.,
International Journal of Chemical Kinetics DOI 10.1002/kin

KINETIC STUDY OF THE CONDENSATION OF SALICYLALDEHYDE WITH DIETHYL MALONATE
597
Table VI Results of Model Fitting: Approximations C, D, and E
Kinetic Model
K-1C
Kinetic Model
4-Piperidinopiperidine
Piperidine
SSR
SD
kr
5 × 10−5
2 × 10−5
7.00 × 10−3
5.00 × 10−3
3.58 × 10−3 15%
6.28 × 10−4 20%
3.47 × 10−4 7.9%
5 × 10−5
6.14 × 10−3 16%
6.39 × 10−4 18%
4.44 × 10−4 11%
2 × 10−5
k
−r
k5
SSR
SD
kr
K-1D
K-1E
7.00 × 10−3
4.00 × 10−3
3.69 × 10−3 36%
6.58 × 10−4 65%
3.50 × 10−4 16%
5.83 × 10−3 2200%
4 × 10−5
9.14 × 10−3 33%
1.86 × 10−3 67%
7.06 × 10−4 31%
8.61 × 10−5 52%
1 × 10−5
k
−r
k5
k7
SSR
SD
kr
6.00 × 10−3
4.00 × 10−3
4.14 × 10−3 29%
2.55 × 10−3 60%
7.56 × 10−4 25%
7.89 × 10−2 5700%
4.17 × 10−5 26%
6.89 × 10−3 16%
1.98 × 10−3 36%
8.50 × 10−4 23%
2.78 × 10−1 950%
3.33 × 10−5 22%
k
−r
k5
k7
k8
[A]0 ꢂ [M]0), the rate of malonate (M) consump-
tion corresponded to the pseudo-first order kinetics:
−d[M]/dt = kobsM[M].
served with respect to aldehyde (A) (see Supporting In-
formation Fig. 4): −d[A]/dt = kobsA. It means that un-
der such conditions there is another rate-limiting step,
and salicylaldehyde takes part neither in this step nor
in the earlier steps. It could suggest that in the excess
of active methylene compound the Hann–Lapworth
mechanism is preferred, but on the basis of additional
experiments under such conditions with Et3N, which
has a similar basicity as piperidine, we did not observe
the condensation reaction. The phenomenon could be
justified taking into consideration formation of sta-
ble enolate—piperidinium complex, as the first step of
the reaction. The decomposition of the complex may
be slower than the next steps and thus determine the
overall condensation rate.
This is in agreement with the proposed mechanism
(K-1C). An excess of salicylaldehyde (A) shifts the
equilibrium to the products side. It could be assumed
that the amine catalyst forms the aminal, and free amine
is not present; in other words, the first stage (i.e., aminal
AP2 formation) is irreversible. Since the concentration
of the aldehyde (A) was approximately constant and the
catalyst regenerated fast, a pseudo-first-order reaction
with respect to malonate (M) was observed.
It should be stressed that the presence of a phenolic
group with essential acidic properties in salicylalde-
hyde (A) and in all the intermediate compounds can
influence the course of the condensation and the effect
must be take into consideration. First, it shifts the equi-
librium of the acid–base reaction to the product side;
the basic catalyst is present in a protonated form. For
that reason, carbanion generation from diethyl mal-
onate is inhibited, and thus formation of β-hydroxy
compound (Scheme 2) is depressed. It was already
found that p-hydroxybenzaldehyde did not react with
various compounds containing an active methylene
group without a base catalyst, because of the acidity of
the hydroxyl phenolic group [10a]. In addition, deam-
ination of β-amino compound (i.e., AP2) (Scheme 3)
can be accelerated in the presence of H+ and both
phenolic group and piperidinium ion can be its source
[7].
CONCLUSIONS
The condensation reaction of salicylaldehyde and
diethylmalonate in the presence of piperidine or
4-piperidinopiperidine in a toluene solution proceeds
via the Knoevenagel mechanism. Taking into consid-
eration the complexity of the investigated reaction,
while based only on kinetic modeling and catalytic
tests, it is hard to predict details of the Knoevenagel
mechanism, i.e., which reaction paths are preferred:
via aminal (which is most probable), iminium ion, or
both. In excess of one of the substrates, additional pro-
cesses might carry out that may strongly influence the
reaction or even may change the reaction mechanism.
In the case of an excess of malonate (M) (i.e., [M]0
ꢂ [A]0), the pseudo-zero-order reaction rate was ob-
International Journal of Chemical Kinetics DOI 10.1002/kin

598
BEDNARZ AND BOGDAL
10. (a) Patai, S.; Zabicky, J.; Israell, Y. J Chem Soc 1960,
2038; (b) Patai, S.; Israell, Y. J Chem Soc 1960, 2020;
(c) Patai, S.; Israell, Y. J Chem Soc 1960, 2025; (d) Patai,
S.; Zabicky, J. J Chem Soc 1960, 2030; (e) Patai, S.;
Edlitz-Pfeffermann, J.; Rozner, Z. J Am Chem Soc 1954,
76, 3446.
BIBLIOGRAPHY
1. Jones, G. Org React 1967, 15, 204.
2. Hann, A. C. O.; Lapworth, A. J Chem Soc 1904, 46, 85.
3. Knoevenagel, E. Ber Dtsch Chem Ges 1898, 32, 2585.
4. Bogdal, D. J Chem Res (S) 1998, 468.
5. (a) Bogdal, D. “Microwaves—Application to solvent-
less phase-transfer catalytic reactions,” Monografia
248, Politechnika Krakowska: Krako´w, Poland, 1999;
(b) Bednarz, Sz.; Bogdal, D. Czasopismo Techniczne
Politechniki Krakowskiej, 2003; p. 2.
6. Kuzmic, P. Anal Biochem 1996, 237, 260.
7. (a) Kinastowski, S.; Mroczyk, W. Bull Pol Acad Sci
Chem 1989, 37, 109; (b) Mroczyk, W.; Szcze˛sna, J.;
11. Terrin, D. D. Dissociation Constants of Organic Bases in
Aqueous Solutions; Butterworths: London, 1965; p. 285.
12. Tanaka, M.; Oota, O.; Hiramatsu, H.; Fujiwara, K. Bull
Chem Soc Jpn 1998, 61, 2473.
13. (a) Lieberman, S. V.; Wagner, E. C. J Org Chem 1949,
14, 1001; (b) Cummings, T. F.; Shelton, J. R. J Org
Chem 1960, 25, 419; (c) Benkovic, S. J.; Benkovic,
P. A.; Comfort, D. R. J Am Chem Soc 1969, 91,
1860.
Kinastowski, S. Roczniki Akademii Rolniczej
w
Poznaniu 1993, 65.
14. Pivonka, D. E.; Empfield, J. R. Appl Spectros 2004, 58,
41.
15. Mroczyk, W.; Szcze˛sna, J.; Kinastowski, S. Roczniki
Akademii Rolniczej w Poznaniu 1990, 57.
8. Kinastowski, S.; Mroczyk, W. Pol J Chem 1984, 58, 179.
9. Tanikaga, R.; Konya, N.; Tamura, T.; Kaji, A. J Chem
Soc, Perkin Trans 1 1987, 825.
International Journal of Chemical Kinetics DOI 10.1002/kin